To determine the effect of proofreading on spelling, a study was conducted with middle- and upper-middle-class students in four sixth-grade classes in two schools in Austin, Texas. Two classes were used as control groups, and two as experimental groups. Teachers of both groups participated in the planning sessions to prepare 14 lessons with materials. Instruction in the use of the dictionary and opportunities for written work were equal in both groups; but instruction in the experimental group emphasized proofreading techniques and involved the use of common-sounds spelling charts and a modification of Dolch's "check-guess" technique. After completing the 14-lesson program, students in both groups wrote stories from which the percent of spelling error was computed for each child's paper. For the experimental group, the time allocated for writing the story included the opportunity to proofread before the assignment was collected. Results indicated that boys who were taught the techniques of check-guessing and proofreading in spelling made significantly fewer errors than those who did not receive such instruction (evidence regarding girls was not conclusive), and that instruction in the use of the dictionary must be specifically directed toward the correction of spelling errors in order to be effective. (SW)